Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Baseball Newsstand > Baseball Primer Newsblog > Discussion
Baseball Primer Newsblog
— The Best News Links from the Baseball Newsstand

Tuesday, March 25, 2014

Miles: Sabermetrics changing baseball’s numbers game

The ribbie rhubarb:

Back when Dusty Baker was manager of the Cubs, he’d go up and down the lineup and say things like Aramis Ramirez was “my RBI guy” and Moises Alou was “my clutch man.”

The whole “clutch hitter” debate is one for another day, but the RBI debate is a fun one to watch today.

Years ago, voters for the Most Valuable Player often tended favor the leader in runs batted in.

Today, thanks again to advanced analytics, the RBI has taken a back seat to stats such as on-base percentage, slugging percentage and OPS, which combines on-base and slugging. OPS-plus, which is scaled to 100 like ERA-plus, is another measure that researchers say gives a clearer picture of offensive prowess.

Other terms, such as “true average,” used by Baseball Prospectus, and “weighted on-base average” (wOBA), used extensively by FanGraphs, now are in vogue.

The problem with relying too heavily on RBI is that the stat depends on the “RBI guy’s” teammates. In other words, if nobody’s on base ahead of a batter who hits a home run, it’s hard for him to rake in the ribbies. But if a cleanup hitter has a couple of .340 or .350 OBP guys batting ahead of him, he’s going to, well, clean up in the RBI department.

In 2005, Cubs No. 3 hitter Derrek Lee had a hitting line of .335/.418/.662 for an OPS of 1.080 and an OPS-plus of 174. He had 199 hits, 50 doubles and 46 home runs.

However, Lee had “only” 107 RBI. You might have figured a guy with Lee’s stats would have been good for 125-130 RBI, but with players such as Neifi Perez (.298 OBP) and Corey Patterson (.254 OBP) hitting ahead of him on many days, it was difficult for Lee to drive up the RBI total.

“This does not make RBIs meaningless, only incomplete,” wrote researcher David Grabiner. “But the real problem with RBIs is ... they measure a lot of things which are not the player’s own contribution. You cannot drive in runners who are not on base (except with home runs), but your own batting doesn’t put them there; if you bat behind good players, you will get a lot of chances.”

Thanks to Butch.

Repoz Posted: March 25, 2014 at 10:26 AM | 29 comment(s) Login to Bookmark
  Tags: sabermetrics

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. Steve Treder Posted: March 25, 2014 at 10:46 AM (#4676710)
And Goldman still rocks.
   2. cardsfanboy Posted: March 25, 2014 at 11:18 AM (#4676728)
Reading the article and he hit one of my favorite ######## points... the blown save in regards to relievers. Technically I don't think there is much they can do about the blown save stat, but I really wish they would list save percentage as blown saves /(saves+holds+blown saves) At least then the percentage would be accurate and you don't get announcers talking about a mid reliever trying for a closers job by talking about their poor save percentage..

Also I think too many people take that Maddux quote
That time coincided with Greg Maddux’ return to the Cubs. I remember chatting with Maddux in spring training and asking him what he thought the most important stat for a pitcher was. He replied: “Wins.”


much too literal. (not the author of this piece who included the second part of the quote) Maddux is ultimately just saying team wins, not even the individual wins.
   3. Mike Emeigh Posted: March 25, 2014 at 02:25 PM (#4676838)
The second part of the quote referenced in #2 is:

Maddux quickly added that he understood stats such as WHIP and ERA, but he felt the pitcher had to do all he could to win the game for his team.


The earlier quote from Roy Halladay is also interesting:

“But I think, ultimately, you look at how guys are able to win games. Sometimes the run support isn’t there, but you sometimes just find ways to win games.”


To me that's the essence of a great pitcher. I've always opposed the idea that a pitcher's performance should be divorced from the context of his run support. I prefer to evaluate starters in terms of how well they perform *against* their run support. If they're getting two runs in a game and their team wins 25% of the time when a typical team would only win about 15% of those games, that's a heck of a performance; if their team only wins 10% of the time under those circumstances, that's not so good.

-- MWE
   4. Lyford Posted: March 25, 2014 at 03:39 PM (#4676866)
"What were the factors that went into the scoring of runs? There were a barrel of them. There was ability to get on base by any means possible. There was power, the ability to hit for extra bases rather than just singles. Then there was speed, daring on the base paths, timeliness of hitting and making the most of opportunities....the ability to get on base, or On Base Average, is both vital and measurable...we found there was still no place for RBIs in the formula. As a statistic, RBIs were not only misleading but dishonest. They depended on managerial control, a hitter's position in the batting order, park dimensions and the success of his teammates in getting on base ahead of him. That left two measurable factors—on base average and power—by which to gauge the over-all offensive worth of an individual..."

Branch Rickey, 1954
   5. Pasta-diving Jeter (jmac66) Posted: March 25, 2014 at 03:50 PM (#4676879)
Branch Rickey, 1954

aka Allan Roth
   6. Moeball Posted: March 25, 2014 at 04:30 PM (#4676905)
Years ago, voters for the Most Valuable Player often tended favor the leader in runs batted in.


Yes, years ago...like in 2012 in the American League. So very long ago.

To me that's the essence of a great pitcher. I've always opposed the idea that a pitcher's performance should be divorced from the context of his run support. I prefer to evaluate starters in terms of how well they perform *against* their run support. If they're getting two runs in a game and their team wins 25% of the time when a typical team would only win about 15% of those games, that's a heck of a performance; if their team only wins 10% of the time under those circumstances, that's not so good.


Mike - some historically interesting tidbits on how pitchers do when getting two runs or less of run support - herein defined as the team scoring two runs or less over the first 9 innings of a game in which the pitcher started. I approached it like this - even knowing a pitcher may have gone only 6 innings or so and that the team run support may not match exactly to how many runs were scored while the starting pitcher was in the game - so that I would not need play-by-play data to make the comparisons. This way I could include pitchers from earlier eras in history. So here's what I found out:

1)From 1962-1966, those fabled last 5 years of Sandy Koufax' career, he went 27-24 when the Dodgers scored two runs or less in his starts. He is the only regular starting pitcher in major league history that I have found to have a winning record over a 5-year period with such poor support. Of course, there could be many factors to help explain this: pitching in an extreme pitchers' park in a pitching-dominant era, maybe Koufax just got an unusually high number of opportunities to pitch in low-scoring games, thus getting a lot more practice at it. Outside of that historic 5-year run, Koufax' record in these situations was only 7-28. On the other hand, his teammate Don Drysdale was pitching under the same conditions as Koufax from 1962-1966 but not getting nearly the same results so maybe there is something to Sandy's performance during his peak years.

While I have not found any other pitchers who matched Koufax' record under these poor-support situations over a 5-year period, there have been some historically great performances under dire circumstances in given seasons, such as the following - not surprisingly, these match with some of the most famous great pitching performances in history:

2)Bob Gibson, 1968 - Cardinals scored 2 runs or less for Gibson in more than half of his 34 starts. He still managed a 10-8 record in these games; he was 12-1 when STL gave him at least 3 runs to work with. At one point early in the season he received 2 runs or less in 8 consecutive starts, the worst unlucky stretch of any pitcher I studied. From June into September, however, he had 6 separate starts where the Cardinals staked him to a 1-0 or 2-0 lead and then he had to hold it because no more runs were forthcoming. In all 6 of these games Gibson pitched a shutout, preserving the slim leads.

3)Gaylord Perry, 1972 - Perry was 24-16 overall that year in winning the AL Cy Young Award. He received 2 runs or less in 25 of his 40 decisions (62.5%), the highest % of bad support of any pitcher in the study. He went 9-16 in these poor support starts. On the other hand, when given at least 3 runs to work with, Perry was a perfect 15-0.

4)Greg Maddux, 1995 - In a higher scoring time in history, Maddux didn't have to worry about getting poor run support such a high % of the time like Perry did. In fact, in 1995 Maddux only had to worry about the Braves being shut out once; he gave up no runs in that start and escaped with a no-decision. He had two 1-0 shutout wins to his credit as well as a 2-0 gem. He had a perfect 3-0 record when given 2 runs or less, which is fantastic! On the other hand, it's a lot easier to achieve that kind of record when you only have 3 or 4 starts all season long where you get such poor support; try doing that when your offense fails to support you almost two thirds of the time like happened to Perry in 1972.

5)Carl Hubbell, 1933 - like Gibson in 1968, managed a 10-8 record when given 2 runs or less. On 7/2 against STL he had a game where he finally got a run in the 18th to win, 1-0. 18 shutout innings pitched to keep the game alive and finally got the win! Followed that up by winning Game 4 in the World Series, 2-1 in 11 innings.

6)The greatest season I've ever seen with poor support: Walter Johnson's historic 1913 campaign. Went 13-5 when given 2 runs or less; in 4 of those losses Washington was shut out (games that no pitcher can possibly win), so Walter actually went 13-1 when given only 1 or 2 runs to work with. That's getting it done!
   7. cardsfanboy Posted: March 25, 2014 at 05:02 PM (#4676934)
Yes, years ago...like in 2012 in the American League. So very long ago.


At least Cabrera had an mvp caliber season, it's not like the Juan Gonzalez over Arod scenario
   8. the Hugh Jorgan returns Posted: March 25, 2014 at 05:38 PM (#4676957)
Yes, years ago...like in 2012 in the American League. So very long ago.


Look, I get this argument and yes I agree that Trout was more valuable, however the first triple crown in 45 years was historically significant and was never going to be denied the hardware. I was actually surprised at how much support Trout received from the voters. And as CFB points out, it's not like it was a slightly above average year, it was MVP caliber. I for one had no issues with them handing the hardware over to Cabrera in 2012, then again I'm nearly 50 so maybe there's some old school bias playing into that.

some historically interesting tidbits on how pitchers do when getting two runs or less of run support


Didn't Blyleven have long stretches where he received minimal run support yet was still "getting it done"
Maybe I'm misremembering...
   9. Moeball Posted: March 25, 2014 at 07:08 PM (#4676998)
Years ago, voters for the Most Valuable Player often tended favor the leader in runs batted in.


Yes, years ago...like in 2012 in the American League. So very long ago.


Look, I get this argument and yes I agree that Trout was more valuable, however the first triple crown in 45 years was historically significant and was never going to be denied the hardware. I was actually surprised at how much support Trout received from the voters. And as CFB points out, it's not like it was a slightly above average year, it was MVP caliber. I for one had no issues with them handing the hardware over to Cabrera in 2012, then again I'm nearly 50 so maybe there's some old school bias playing into that.

Yes, and part of Cabrera winning the traditional Triple Crown was leading the league in RBIs - had he not done that, he wouldn't have won the Triple Crown, so of course, he therefore wouldn't have won the MVP when Trout clearly had the superior season...oh, wait, now I've just described the 2013 season.

Never mind.
   10. cardsfanboy Posted: March 25, 2014 at 08:26 PM (#4677020)
Yes, and part of Cabrera winning the traditional Triple Crown was leading the league in RBIs - had he not done that, he wouldn't have won the Triple Crown, so of course, he therefore wouldn't have won the MVP when Trout clearly had the superior season...oh, wait, now I've just described the 2013 season.


Trout's clearly superior season in 2012 is based upon 1. Park effects and 2. defense ....as shown by his 2013 his defense is probably overrated in 2012 by the war numbers, and it's arguable that there is something screwy going on with AL west park effects.

But if you look at the numbers, it's not a "clearly" superior season, Trout is a superior season when you look deaper, but not clearly.

Cabrera had the better avg, slg and ops. Played in 19 more games/60 more pa, Was on a pennant winning team, moved positions to accomodate a new acquisition etc...Considering that the park effects for Detroit have gone from neutral to a hitters park and anaheim has gone from neutral park to a very good pitchers park over the past few years, you can maybe understand why people aren't so willing to automatically accept the changing park factors as gospel.

Those who insist that War is the only benchmark for MVP is missing a big point.
   11. haggard Posted: March 25, 2014 at 08:56 PM (#4677026)
I'm pretty sure mvp voters of the past were aware that rbis are a team achievement. I'm also pretty sure that's the main reason why they gave some of those mvp awards to rbi guys.
   12. BDC Posted: March 25, 2014 at 10:50 PM (#4677054)
One thing's for sure: Runs Scored, as team-dependent as RBI, and as important to winning or a bit more so (not every run is batted in), haven't been seen as indicating value as much as RBI have, and still aren't. Trout after all has two Runs titles which have drawn yawns from old-schoolers & statheads alike.
   13. Baldrick Posted: March 25, 2014 at 11:33 PM (#4677062)
Cabrera had the better avg, slg and ops. Played in 19 more games/60 more pa, Was on a pennant winning team, moved positions to accomodate a new acquisition etc.

Moving positions to play a mediocre/poor third base is nice and all. But you don't need park effects and excellent defensive ratings to think Trout was clearly better. You need those to think that he was four wins better. Even with extremely weak assumptions about defense and park, Trout was obviously the better player. Cabrera won because people still care a lot about the triple crown numbers, and because his team made the playoffs. Both are pretty dumb reasons, which is an argument against the change discussed in the article.

It's not like Cabrera's two MVPs these years are equivalent to Juan Gonzalez. But there's no denying that fascination with less-relevant numbers had a huge part in them.
   14. Squash Posted: March 26, 2014 at 12:36 AM (#4677082)
To me that's the essence of a great pitcher. I've always opposed the idea that a pitcher's performance should be divorced from the context of his run support. I prefer to evaluate starters in terms of how well they perform *against* their run support. If they're getting two runs in a game and their team wins 25% of the time when a typical team would only win about 15% of those games, that's a heck of a performance; if their team only wins 10% of the time under those circumstances, that's not so good.

The problem with this is that since there are relatively few games in which pitchers pitch where their team scores fewer than two runs you're baking a ton of variability into the pie (to reuse a phrase I used in a different thread a few days ago). I imagine we don't really care how guys do when they're given four or five runs a game - those are games pretty much every pitcher "should" win, as if you have an ERA in the 5 range you're not going to be starting pitcher for very long - hence we're left with the low-scoring games to "prove" who the great pitchers are. Do you really want to be judging how great a pitcher is in his career by 40 or 50 starts?
   15. CFiJ Posted: March 26, 2014 at 01:38 AM (#4677091)
In 2005, Cubs No. 3 hitter Derrek Lee had a hitting line of .335/.418/.662 for an OPS of 1.080 and an OPS-plus of 174. He had 199 hits, 50 doubles and 46 home runs.


Ah, Deracles. *sniff* *wipes tear from eye*
   16. Rob_Wood Posted: March 26, 2014 at 03:40 AM (#4677098)

To follow up on MikeE's perspective (and Squash's comment), I developed a stat I called "Win Values" many years ago that does exactly that for every start a pitcher has in a season. For every possible run support a pitcher could receive, I calculate the team's expected winning percentage for that game. Then, on a game-by-game basis, I derive how much the starting pitcher's performance increased or decreased his team's chance of winning that game given his runs given up.

Of course, a pitcher who throws a shutout in a 1-0 game receives a lot of the credit for the team win since the team had a very low likelihood of winning a game while scoring only one run. On the flip side, a pitcher who throws a shutout in a 9-0 game receives virtually no credit for the team win since the team had a very high likelihood of winning a game while scoring nine runs.

Anyway, I posted the methodology and the results in a series of articles years ago.
   17. cardsfanboy Posted: March 26, 2014 at 07:52 AM (#4677115)
Moving positions to play a mediocre/poor third base is nice and all. But you don't need park effects and excellent defensive ratings to think Trout was clearly better. You need those to think that he was four wins better. Even with extremely weak assumptions about defense and park, Trout was obviously the better player. Cabrera won because people still care a lot about the triple crown numbers, and because his team made the playoffs. Both are pretty dumb reasons, which is an argument against the change discussed in the article.


Trout was clearly better on a per game basis... not disputing that at all, but it's ridiculous to assume he was clearly better on a cumulative basis. 20 games is a lot of missed action. Gun to my head, I go with Trout, gun to my head I also claim 4 wins is equally as ridiculous assumption as arguing for Cabrera over Trout on a rate basis.


It's not like Cabrera's two MVPs these years are equivalent to Juan Gonzalez. But there's no denying that fascination with less-relevant numbers had a huge part in them.


Agreed, I'm not arguing for Cabrera over Trout. I'm arguing that it's defensible position without requiring a cult like worship of RBI. (Just like I think the cult like worship of War for Trout probably overstates Trout's value in 2012)
   18. Rickey! On a blog from 1998. With the candlestick. Posted: March 26, 2014 at 10:14 AM (#4677187)
Hello. My name is Sam and I'm a test case.
   19. Hal Chase School of Professionalism Posted: March 26, 2014 at 01:21 PM (#4677302)
Went 13-5 when given 2 runs or less; in 4 of those losses Washington was shut out (games that no pitcher can possibly win), so Walter actually went 13-1 when given only 1 or 2 runs to work with. That's getting it done!


What was Walter doing the three times he came to bat in each of those games they were shut out? He was probably b*^%ing about run support.

Get a hit, Walter. Turn the lineup over, you sandbaggin' sonofabitch.*


*I am, of course, tongue-in-cheek here. Just being a pain and pointing out that there was one way The Big Train could have guaranteed himself a win.
   20. Barry`s_Lazy_Boy Posted: March 26, 2014 at 02:16 PM (#4677349)
Runs Scored, as team-dependent as RBI

Strongly disagree.
   21. cardsfanboy Posted: March 26, 2014 at 02:48 PM (#4677381)
Strongly disagree.


Agree. It's team dependent, but not nearly to the same degree as rbi.
   22. SoSHially Unacceptable Posted: March 26, 2014 at 02:55 PM (#4677388)
Agree. It's team dependent, but not nearly to the same degree as rbi.


There may be a slight difference, but they're pretty damn close in terms of team dependency.
   23. Karl from NY Posted: March 26, 2014 at 04:36 PM (#4677486)
How are runs not exactly as team-dependent as RBI, aside from the occasional wild pitch or some such? Anyone on base needs someone to drive him in, and anyone driving in a run needs somebody on base to drive in. And you can do either by yourself with and only with a HR.
   24. cardsfanboy Posted: March 26, 2014 at 05:15 PM (#4677507)
How are runs not exactly as team-dependent as RBI, aside from the occasional wild pitch or some such? Anyone on base needs someone to drive him in, and anyone driving in a run needs somebody on base to drive in. And you can do either by yourself with and only with a HR.


The difference really is just odds. If you get on base a lot, even a poor team will drive you in from time to time. You are only dependent on really one thing happening not in your control. Add in speed, and that can help your odds some(also getting yourself in scoring position etc.)

With RBI you are dealing with having someone get on base in front of you and you timing your random positive offense with them. Or if you are named Votto or Bonds, not really getting the chance to produce as the opposing team works around you.

Other words...every time you get on base you have a chance of scoring.... with rbi, outside of a homerun, you are going to be limited based upon your team getting someone on base in front of you. They are both of course team dependent, but with runs, it's your skill at getting on base/in scoring position that is the biggest factor, while rbi is limited to the luck of how often your teammates show their skill.


Edit: I'm not sure how that reads...hopefully it's clear.
   25. SoSHially Unacceptable Posted: March 26, 2014 at 05:39 PM (#4677529)
Other words...every time you get on base you have a chance of scoring.... with rbi, outside of a homerun, you are going to be limited based upon your team getting someone on base in front of you. They are both of course team dependent, but with runs, it's your skill at getting on base/in scoring position that is the biggest factor, while rbi is limited to the luck of how often your teammates show their skill.


Nah, that's not it. Because each time you get on base, save home runs, you need someone else to move you around in one way or another. In that sense, it's the exact same thing.

Each run is made up of four bases. The most important of those bases is the first one, which is why run scored is more individual-dependent than team. But the run scorer is still typically relying on teammates, sometimes more than one, to move him around the bases, just as the RBI guy is dependent on guys getting on base in front of him.

And just as RBIs can be produced on outs (sac flies, groundballs), runs scored can be as well (batter reaches on a forceout or other FC). Moreover, in some cases, the guy scoring the run didn't come to bat at all (a pinch-runner), a situation that doesn't exist for the RBI guy.

Altogether, I'd say the RBI is slightly more team-dependent than the run scored, but the difference is tiny.
   26. Karl from NY Posted: March 26, 2014 at 05:47 PM (#4677535)
You are only dependent on really one thing happening not in your control. ... With RBI you are dealing with having someone get on base in front of you and you timing your random positive offense with them.


These are the same thing. The run-scorer is just as dependent on having the RBI guy randomly timed for a hit behind him as the RBI guy is dependent on the run-scorer randomly timed to reach base ahead of him.
   27. Sunday silence Posted: March 27, 2014 at 06:17 PM (#4678015)

I've always opposed the idea that a pitcher's performance should be divorced from the context of his run support. I prefer to evaluate starters in terms of how well they perform *against* their run support


Mike: I found your comment very interesting, do you also prefer to evaluate starters in terms of the actual game conditions under which they pitched?

To be more specific; since it is thought that park effects are significantly effected by weather, then is it possible to further refine a pitchers ERA+ to take account those particular conditions that he pitched on particular days? AN obvious example would be a pitcher who makes say 5 starts in Wrigly when the wind is blowing out, he has an ERA of 5 in those games but that could be quite brilliant given the weather on that particular day.

Would you be in favor of such a method? or not enuf sample size?
   28. PreservedFish Posted: March 27, 2014 at 06:53 PM (#4678036)
I've always opposed the idea that a pitcher's performance should be divorced from the context of his run support. I prefer to evaluate starters in terms of how well they perform *against* their run support


Has anyone taken a stab at this stat? Obviously it would need to go inning by inning, so you don't give a guy credit for pitching to the score when what really happened is that he got hit hard and his team ended up bailing him out.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
Jim Wisinski
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Hot Topics

NewsblogMLB: Tarp problems at Yankee Stadium
(8 - 7:58am, Jul 24)
Last: RMc's desperate, often sordid world

NewsblogCSN: Enough is enough — time to move on from Ryan Howard
(42 - 7:57am, Jul 24)
Last: PASTE Thinks This Trout Kid Might Be OK (Zeth)

NewsblogOTP - July 2014: Republicans Lose To Democrats For Sixth Straight Year In Congressional Baseball Game
(2871 - 7:51am, Jul 24)
Last: Bitter Mouse

NewsblogAs shifts suppress offense, time has come to consider a rule change
(78 - 7:45am, Jul 24)
Last: Jeltzandini

NewsblogGeorge "The Animal" Steele Mangles A Baseball
(131 - 7:44am, Jul 24)
Last: a fatty cow that need two seats (cough, cough)

NewsblogOT: The Soccer Thread July, 2014
(369 - 7:03am, Jul 24)
Last: Swedish Chef

SABR - BBTF ChapterWho's going to SABR??
(86 - 5:36am, Jul 24)
Last: McCoy

NewsblogOT: Monthly NBA Thread- July 2014
(852 - 3:31am, Jul 24)
Last: MikeOberly

NewsblogSports Reference Blog: 1901-02 Orioles Removed from Yankees History
(30 - 2:39am, Jul 24)
Last: DFA

Newsblog2015 Competitive Balance Lottery Results
(9 - 2:29am, Jul 24)
Last: DFA

NewsblogOMNICHATTER 7-23-2014
(50 - 2:16am, Jul 24)
Last: Crispix reaches boiling point with lackluster play

NewsblogGoldman: Eliminating the shift a bandage for a phantom wound
(15 - 1:28am, Jul 24)
Last: jwb

NewsblogRubin: deGrom for NL rookie of the year?
(36 - 12:28am, Jul 24)
Last: KT's Pot Arb

NewsblogCameron Maybin Suspended 25 Games
(17 - 11:36pm, Jul 23)
Last: Nasty Nate

NewsblogNats Fan in Colorado Shows Support with Curly W BEARD
(2 - 10:29pm, Jul 23)
Last: boteman is not here 'til October

Page rendered in 0.2371 seconds
52 querie(s) executed